Review of Commercial Electronic Ceramic Materials in

Australia; 50 years onwards


Gerry Triani1  and Peter Bryant2


1) Institute of Materials Engineering, Australian Nuclear Science and Technology Organization (ANSTO), Lucas Heights, NSW 2234

2) Thales Australia, Rydalmere NSW 2116


Email: and


Available Online at:




This review paper will address the Australian history of electronic ceramic materials which were used primarily for their dielectric properties rather than their insulation properties. The history of these materials can be broken down into three main periods. The first period is that prior to 1950, when the electronic components industry was

first set-up. The second period from approximately 1950 to 1975 involves the expansion and decline of the capacitor product lines and production capacity under the Ducon Condenser Company. The third period from approximately 1975 and continuing to the current day, involves the manufacture of piezoelectric components and

devices through Ausonics Pty Ltd and a group at Plessey Ducon, which was subsequently managed in turn by its successors, Plessey Australia, GEC Marconi, Thomson Marconi and the Thales Australia companies.




This early period is undocumented in the current

Thales Australia technology database. A brief history can be gleaned from historical newspaper records.


Following the commencement of commercial radio broadcasting in Australia in 1922, the radio found increasing acceptance in households in Australia and many companies were setup to manufacture radios  and  the  components  that  were  used  to support this manufacturing. Companies such as Nilcrom Porcelains Pty. Ltd. and  Ferro Enamels Ltd.  developed  locally  manufactured  insulators from porcelains and steatite [1].


Ducon Condenser Pty. Ltd. was another such company started in 1932 to supply capacitors into this market and it quickly grew to dominate the Australian market [2]. The company was founded by a German immigrant, Hans? Hecht who after a chance conversation with Andrew Perrson, a Swedish immigrant, decided to invest £5000 in this new venture. This gave Hecht 50% ownership in the new company [3]. Hecht had wide ranging business interests involving a number of import- export companies and for a while in the mid 1930s was involved with the Jubilee radio. [4] Hecht claimed that the growth of Ducon was largely due to the efforts of Perrson and also Cliff Gittoes who later became managing director and chairman [3].

The expansion of Ducon and the other companies was furthered by the need to develop a self- sufficient Australian industrial capability during World War II. In some cases, this involved identifying and utilising Australian raw materials. [1].  In  other  cases, Ducon took  over companies with specific product ranges such as Simplex Products Pty Ltd which made mica based capacitors [5] and where significant industrial shortcomings were identified, Ducon undertook technology transfer arrangements with companies such as Compagnie nérale De La Télégraphie Sans Fil of France, General Ceramics Corp, Erie Resistor and P. R Mallory Inc. of the USA [6, 7].


In this period, Ducon developed some capability with  manufacture  of  ceramic  capacitors.  People such as J.M. Gleeson of General Ceramics and Steatite Corp. helped set up the ceramic facility in the 1940s with particular emphasis on the steatite (magnesium silicate) based materials. The low dielectric losses and high dielectric and mechanical breakdown  strengths  of  such  ceramics  found  a ready market in many communications systems [8]. Unfortunately, there are no detailed records of the materials used and manufactured at that time.



FROM 1950 TO 1975

The technology and manufacturing developments

of this period are well documented by internal company reports written by Hans Ruckert. These



reports are currently held in the Materials Laboratory of Thales Australia at Rydalmere. Hans Ruckert was a German scientist who had worked in the HF Dielectric Laboratories of Rosenthal- Insulator GmbH. Barium titanate was discovered and developed independently in the USA, Russia and Japan [9]. It quickly became the key dielectric material as it allowed the manufacture of products with  a  wide  range  of  dielectric  constants, dissipation  factors  and  temperature  coefficients. The management at Ducon recognised the need to make   products  based   on   barium   titanate  and initially sought to procure the expertise through technology transfers with the Erie Resistor company. These efforts were not particularly successful because the American personnel were only employed on short term contracts and secondly, because the manufacturing equipment used (e.g. oil fired furnaces) were poorly suited to advanced ceramic manufacture.


Australia had established a Ministry for Industrial Development in the late 1940s to make the country more independent from imports and so the migration of people with special experience was organised. Hans Ruckert was one of these people. As Hans himself noted, he had no precise formulation experience but by the end of 1951 had established the compositions and processes to manufacture the N750, NPO and K4000 materials which could be used for both tubular and disc capacitors. Over the next 20 years the list of dielectric  materials  was  improved  and  expanded and even today some of his NPO formulations have not been bettered [10].


It is instructive to note many of the issues that Hans

Ruckert faced in the development of such materials.

 A failure of management to quickly appreciate that a high performance product requires more stringent manufacturing conditions

 Incomplete transfer of information during a technology transfer

 The possible over-reliance on a single person as the source of the technology

 The need to produce materials to international standards (IEC or EIA classifications)

 Inadequate test equipment

 Oil      fired     furnaces    producing   reducing atmospheres rather than oxidising

 Inadequate furnace materials e.g. zircon setter rather than the less reactive stabilised zirconia

 Inadequate raw materials

 The  need  for  cleanliness  and  a  dust  free environment

 The need for a superior electroding material


Component production and engineering activities reached a high standard in the late 1960s and 1970s but the component designs remained at the discrete level. Electronic equipment designs had also stagnated and left the television and radio manufacturing industries very vulnerable to competition. When Japanese companies were allowed to establish television manufacturing plants in Australia in the early 1970s, they imported completely  assembled  circuit  boards.  The Australian companies were forced to do the same thing and with the tariff reductions in 1973, the Australian component industries which included capacitors, resistors, semiconductors, ferrites and volume controls etc. were all decimated [10].



Thousands of jobs and much experience were lost in this period. Yet in hindsight, one might also reflect that given the ways in which the electronics components industry has developed over the last forty years with devices such a multilayer chip capacitors and integrated circuits that Australian industry was in a poor position to be at the forefront of such developments. One might also pinpoint a tendency   for   the   companies   to   be   primarily focussed on just the static Australian market, to have a restricted research and development group centred on one person who focussed on materials rather than one with a multi-disciplinary team with a focus on product applications and efficient material processing, and the existence of government policy which effectively destroyed an industry rather than encouraging its evolution and long term survival [11].






Ducon manufactured a full range of dielectric materials  and  produced  materials  to  the International Electrotechnical Commission (IEC) standards of the time. A similar classification applies today where IEC/EN 60384 series is the appropriate standard. This document series defines the class 1 (the temperature stable materials) and class  2  (high  dielectric  constant)  ceramic dielectrics. The old class 3, barrier layer dielectrics, is now considered obsolete but was a competitive product in the late 1960s and early 1970s.


The class 1 dielectrics based on paraelectric materials were defined by both their dielectric constant or “K value and their temperature coefficient of capacitance. The temperature coefficients were designated in accordance with the IEC standard as P100 or N750 where the P or “N designated either a positive or negative temperature coefficient in ppm/K respectively. The temperature stable materials were designated as NPO.


The starting points for development of the type I ceramic dielectrics range were the K6.5/P150 steatite and the K90/N750 titanium dioxide base ceramics. The importance of using a fully oxidising atmosphere and the usefulness of additives such as manganese dioxide in controlling the dissipation factor were key early developments. Subsequently, in line with other manufacturers, medium K/TCC dielectrics based on barium titanate + titanium dioxide were produced and commercialised.

Ideally, one would prefer a temperature stable material but there was usually some trade-off between  temperature  stability  and  dielectric constant and a larger dielectric constant allowed a particular component to be physically smaller. In many  instances,  the  negative  temperature coefficient of capacitance was selected to balance positive coefficients of other circuit components such as ferrites in tuned circuits so that the overall circuit stability was improved. For the NPO materials, a range of dielectric constants was available. For leaded components with low capacitance values, it was preferable to use a lower dielectric  constant  material  to  enable  easier handling and soldering. For very large high-voltage capacitors, a lower dielectric constant was often preferred  as  the  breakdown  voltage  of  the  unit could then be optimised.


The K105 NPO was a development of the early

1970s. This represented a significant advance over previous best K values in the 60-80 range. A draft

patent for this material had been prepared and discussions on licensing to foreign firms had also commenced  at  the  time  the  dielectric  ceramic

manufacturing plants were shut down and so never saw any significant production. Even today, this material would represent state of the art in such

temperature stable ceramic dielectrics. Modifications to this NPO composition allowed the production  of   novel  composition  such  as   the

K200/N1000 material.


A listing of the more frequently produced materials and   their   base   compositions  is   given   in   the following table [12].




Table 1. The Plessey Ducon range of Class1 dielectric materials and their key properties


Ceramic names

Temperature coefficient α

α-Tolerance ppm /K

IEC/ EN- letter-code


Main compositional components


P150                                  150                                         ±50                                      --                       magnesium silicate

P100                    100                         ±30                     AG             TiO2+ZrO2+SnO2

NP0                      0                           ±30                     CG             K16: magnesium titanate

K37: barium titanate + TiO2

K105: barium titanate + rare earth titanate

N33                    33                         ±30                     HG             Barium titanate + TiO2

N75                    75                         ±30                      LG             Barium titanate + TiO2

N150                  −150                        ±60                      PH             Barium titanate + TiO2

N220                  −220                        ±60                     RH             Barium titanate + TiO2

N330                  −330                        ±60                      SH             Barium titanate + TiO2

N470                   -470                        ±60                      TH             Barium titanate + TiO2

N750                  750                      ±120                     UJ             K90:TiO2subscripts?

K160: barium bismuth calcium titanate

N1500                1500                      ±250                     VK             calcium titanate + zirconate

N3300                -3300                         --                         --              barium titanate + strontium zirconate titanate

N7500                -7500                         --                         --              barium titanate + strontium titanate



Ducon’s range of class 2 dielectrics was based on ferroelectric compositions and was quite conventional. A range of barium titanates doped with Curie Point shifters such as calcium zirconate and strontium titanate was employed. Again, significant use was made of manganese dioxide to better control the dissipation factor. Up until the

1970s, Ducon classified all its type II dielectrics purely in terms of their dielectric constants. In more

recent years, the IEC have defined the temperature coefficients  of  capacitance  more  precisely. Although a very large number of possible compositions are allowed under the standard, most

production seems to comply with the following three categories.


2X1   Temperature   range      5C   to   +125°C;

variation +/- 15%

2E6  Temperature range +10°C to +85°C; variation


2F4 Temperature range 2C to +85°C; variation



Although the variation of capacitance with temperature was measured for the Ducon compositions, it  was  not  reported  in  accordance with the latest IEC standards. The following table lists  the  various  Ducon  grades  produced  in  the

1970s along with the reported temperature variations and possible allocations to the IEC equivalents. It is the authors’ opinion that these Ducon materials may not be fully compliant to the IEC standards particularly at the temperature extremes as such requirements were not design criteria when the materials were developed.


It is also interesting to note that in practical terms, the Ducon materials were developed so as to be blendable, so that adjacent pairs could be mixed so as to generate intermediate values of the dielectric constant if required. Shortly prior to plant closure, this concept was to be extended so that the production department would only need to make the two end members of this series and any intermediate members would be obtained by blending [13].


The Class 3 dielectric materials used for what are commonly called barrier layer capacitors” are obsolete in the latest IEC standards as they have effectively been replaced by the higher dielectric constant class 2 dielectric in multilayer packages. The class 3 dielectrics were originally purported to have dielectric constants well in excess of 100,000. Such high dielectric constants enabled the production of high capacitance values in small package sizes. The drawback to these components was their low insulation resistance and low working voltages.


Hans   Ruckert   and   his   team   at   the   Ducon

capacitor” was in many ways a misnomer. They were able to identify a number of key factors which enabled the production of such devices with high capacitance, insulation resistance and good working voltages. These devices were better described as reduced core ceramic capacitors.


The key parameters include


i)            A starting composition comprising a barium titanate with an excess of TiO2 and a rare earth titanate with is readily fired in an oxidising atmosphere to give a ceramic with a dielectric constant of ~6000 at 25°C and a tan δ of ~ 1.5%

ii)           The sintered ceramic should have a fine uniform grain structure. Contamination  and  damage  to  the parts must be avoided. Use of raw materials with very low alkali metal components as the presence of such compounds can prevent the proper re- oxidation of the outer layer.

iii)          A reduction firing which reduces the ceramic   parts   fully   through   their

thickness. This was typically done by

firing the parts stacked on their sides in  inconel  boxes  while  sitting  on  a

layer of carbon black.

iv)          An  acid etching process to  increase the     surface    area     followed    by

impregnation   with     a     manganese

nitrate solution.

v)           A   re-oxidation   firing   which   re- oxidises the skin and subsequently fully  separates  the  metal  electrodes and the reduced core. The skin layer must be greater than 2.5 microns (or at least two grain layers) otherwise the reduced core will be partially exposed and the device will remain semi- conductive. The temperature and time of    the    re-oxidation    firing    was precisely  controlled  to  obtain  a sharply defined oxidised skin with a high insulation resistance.

vi)          Green  density  variations  within  the part must be minimised

vii)         The     re-oxidation     must     proceed

uniformly on both major surfaces

viii)       When using fired on electrodes, the glaze type and amount must be optimised. The glaze oxidises a reduced ceramic skin and then further insulates the skin from further oxidation  preventing  it  from becoming too thick.

ix)           The  electrode  firing  cycle  must  be optimised and each face of the part must   be   exposed   to   an   identical




Table 2: The Plessey Ducon range of Class 2 dielectric materials and their key properties


Ceramic name



Temperature for

maximum KT33






- X%

Lower temperature for -X% °C

Higher temperature for -X% °C

Possible IEC grade

K2400            2400                      25                     5%                      0                               98                        2X1

K4500            4500                      25                    10%                    -15                             50                     2E6/2F4

K6500            6500                      25                    20%                    10                              38                     2E6/2F4

K10000          10000                      23                    50%                    -20                             57                        2F4

K14000          14000                      20                    50%                     7                               57                        2F4

K15700                     15700                                      23                                 50%                                  -12                                               45                                       2F4


Ducon made millions of these types of capacitors which it called red caps for the transistor radio market. Working voltages from 3V to 100V were achieved by adjusting the re-oxidising process to increase the skin thickness from ~2.5 microns to ~

100 microns. As noted above the skin layer could not be made thinner than 2.5 microns without the reliability of the device being affected [10, 14].




With the closure of the Plessey Ducon ceramic dielectric department in Villawood NSW, much of the   skills   and   knowledge   was   virtually   lost.

However, a small group was retained to work on piezoelectric materials, primarily for projects for the Department of Defence. One of these programs

involving the Barra sonobuoy was the primary driver ensuring the survival of electronic ceramic manufacture  to  the  current  day.  Some  of  the

equipment   was   moved   to   the   main   Plessey

Australia site at Meadowbank NSW and remained there until 1996 when the facility was transferred to

a  new  site  at  Rydalmere  operated  by  Thomson

Marconi sonar.


Ducon had made a piezoelectric barium titanate for a number of years in the 1960s. In addition many of the  class  2  dielectrics  would  have  been piezoelectric below their Curie Points but as this temperature had been deliberately adjusted to around 2C, such materials had no practical application as piezoelectrics. The piezoelectric composition used was compliant to the Type IV requirements as defined by MIL_STD_1376 (SHIPS) 1970 Piezoelectric Ceramic for Sonar Transducers.


The situation changed in 1967 when DSTO expressed as interest in using PZT based compositions. Hans Ruckert and his team then commenced a development program focussed on PZT ceramics. For the first few years, considerable effort was expended in reviewing the literature regarding the compositions and processing of PZT materials. In addition to this, experience was collected on PZT processing and in particular the calcining and sintering of a material with a highly volatile ingredient. The lead oxide component of

PZT bodies is so volatile above 900°C that a useful component cannot be manufactured unless a practical way is found to control its volatilisation [15].


It was also realised that the principal commercial compositions were those produced and licensed by the  Brush  Clevite  Corp  (subsequently  Vernitron and then Morgan Electro Ceramics) throughout the USA and Europe. The importance of the Brush Clevite materials was further emphasised by the release  of  the  US  Military  standard MIL_STD_1376 (SHIPS) 1970 Piezoelectric Ceramic for Sonar Transducers in December 1970. This document quickly became a de facto standard for the piezoelectric industry [16].


The situation in Japan and Russia was different. In Japan, Ouchi and his co-workers at Matsushita developed materials based on the lead magnesium niobate-PZT system. [17]Ref The Russians also developed a  range of  very similar compositions. The Ducon team realised that the lead nickel niobate-PZT system offered similar, if not superior performance and ease of processing to the Matsushita system, and Hans Ruckert was awarded a patent for this compositional family in 1970 [18].


This patent enabled a continuum of soft PZT materials to be produced with dielectric constants ranging from ~ 1500 to ~ 6000 by simply varying the amount of lead nickel niobate. This exceeded the range defined in MIL-STD 1376.  Use of nickel rather than magnesium, as in the Matsushita material, enabled the manufacture of a material which  more  consistently  could  be  made  in  the active perovskite form. The patent also defined manganese dioxide (or similar) as the critical additive for converting the soft PNN-PZT materials into harder materials suitable for active transducers. An appropriate selection of the amount of lead nickel niobate and the amount of manganese allowed compositions virtually equivalent to those defined in MIL_STD_1376 (SHIPS) 1970.

With this understanding and patent protection, Plessey Ducon began manufacturing PZT ceramics in 1969. Major early markets for these materials

included the Sonalert transducer for telephones for

the  then  PMG  (now  Telstra)  and  discs  for  the



Ausonics Pty Ltd Octoson medical ultrasonic scanner. The principal market and the driving force which ensured the survivability of the ceramics manufacturing capability was the Barra hydrophone and sonobuoy developed in conjunction with AWA and DSTO.


Hans Ruckert retired in 1979 and Technical and Business reins were taken up by Dr Gavin Tulloch. Over the next 10 years Gavin Tulloch greatly expanded the product range and technical skills of the  department  so  that  it  became  an  integrated Sonar Products house with both design and manufacturing capability for both piezoelectric components and sonar transducers. The products manufactured in this Plessey Australia period included the Mulloka active transducers arrays, the Jezebel hydrophones, the active and passive transducers on the Oberon class submarines and the Kariwarra hydrophones and towed arrays.


In the GEC Marconi period from 1988 onwards, the growth of the facility stabilised and the department was involved in the supply of transducers for nearly all  the  major  sonar  platforms  of  the  Australian navy. These included the Scylla Sonar on the Collins  Class  submarine,  the  2093  Minehunter sonar on the Huon Class minehunters, the EMATT expendable  sonar  training  targets,  the  Spherrion hull mounted sonar, and the Coltas/Shortas towed array sonars.


From the early 1990s to 2006, the company was involved in the design and manufacture of slimline towed array sonar systems for the seismic industry. The department was then involved with the design and manufacture of over a million hydrophones into this market over a 10 years period. Since the end of this program the department has again concentrated on supplying the military market and has also exported a significant amount of ceramic and transducers to sister companies in Europe. This activity is ongoing and the export represents the majority of the current business.


Another commercial activity utilising piezoelectric ceramics in the mid 1970s was the medical diagnostics company Ausonics Pty Ltd located at Lane Cove in Sydney. This company emerged to develop and manufacture an Australian innovation from the Ultrasonics Institute, then part of the Commonwealth  Health  Department,  later transferred to CSIRO. George Kossoff, David Robinson and colleagues pioneered the use of ultrasonic medical diagnosis using grey scale ultrasound’ for obstetrics imaging utilising a  2.5

MHz transducer fitted with piezoelectric ceramic

elements [19]. Ausonics Pty  Ltd  commercialised the UI Octoson technology in 1976, developing hospital systems where the patient would lie on a water bed covered with a flexible membrane. The ceramics for this application were originally manufactured by Plessey. Pulse-echo responses were transmitted through the water providing images of the foetus inside the uterus. In the mid

1980s, Ausonics developed real-time’ diagnostic equipment  using  mechanical  section  scanner  to

generate ultrasonic images of the body. Although

robust for their application, piezoelectric ceramics of   specific   frequency   (i.e.   2.5-7.5MHz)   were

required  to  drive  the  transducer  around  a  pivot point at the front of the scanner. This was achieved by machining a large block of ceramic to a specific

radius of curvature and thickness to achieve the desired frequency range. However, by the 1990s the trend in the diagnostics industry was to move away

from mechanical systems to arrays, which allowed the transducer beam to steer electronically, thus avoiding moving parts.


In the late 1980s the Australian Nuclear Science and Technology Organisation (ANSTO) applied its nuclear ceramic expertise to the field of piezoelectric ceramics. ANSTO conducted a 3 year project with Ausonics focused on developing the next generation piezoelectric ultrasonic transducers using composite (ceramic/polymer) structures [20]. This project involved advanced powder processing suitable for tape-casting as a means to produce cost effective near net shaped ultrasonics devices.


Tape casting has become an established ceramic processing method for forming, thin, flat, large area ceramics in the thickness range 10-3000 µm with high precision and is used to fabricate ferrite memories, ceramic capacitors and electrically insulating substrates for thick and thin film multilayer circuitry [21]. The tape-casting process involves the suspension of finely divided ceramic powders  in  an  aqueous  or  organic  vehicle comprised on solvents, deflocculants, plasticisers and  binders  to  form  slurry  that  is  cast  onto  a moving carrier tape. The slurry passes beneath the knife-edge of a blade that levels the slurry to form a layer  of  controlled  thickness  and  width  as  the carrier surface advances along a supporting table (Figures 2a and b). When the solvents have evaporated, the fine solid particles coalesce into a quite dense flexible sheet which is essentially a collection  of  particles  bonded  by  the  polymer phase. The flexible green ceramic” sheet may be stored on take-up reels or stripped from the carrier, into continuous rolls.




Fig. 2: (a) Doctor blading process used to produce ceramic tape and (b) Ceramic cast tape drying on a polymer


carrier to produce large surface area PZT material for shaping and sintering. Typical wet cast gap 1.0 mm


























Fig. 3: PZT tape (bottom and right hand elements) and sintered net-shaped curved discs. A curved piezoelectric ceramic disc is embedded into the mechanical sector transducer (left). A linear array with 20 elements aligned upright is also shown (centre).



Table 3: Piezoelectric ceramic specifications of ANSTO formulation in comparison to commercial product


Property                         Vernitron 7A                     Toshiba C-24              ANSTO formulation

Q                                        600                                       920                                      <920

ε33/ε0                                                                                                                  425                                       209                                   100-250

Kp                                                                                                                         0.51                                                             0.04                                                           <0.04

K33                                                                                                                        0.5                                       0.52                                   >0.52 d33  10-12 m V-1      150          68.2                                                      >68.2 g33   10-3 V.m N-1                                  39.9                                      36.8         37-50

Z ( 106 kg m.s-1)                              36                                         32                                      32-36


Curie temp  (ºC)                           350                                       255                                   200-350



A  key  challenge  of  the  composite  piezoelectric project was to sinter tape ceramic to high density

strictly fails to meet the aging rate criteria of the


MIL-STD 1376B (its aging rates are too low) and

final      dimensions.     Linear      shrinkage     was

the percentage change in KT

is too high. The d33

approximately between 14-20% depending on the

solids loading but could be controlled to +0.5%. Alumina   substrates   fabricated   for   production

purposes had tolerances for thickness of + 10%, and flatness (camber) of + 0.4%, so one of the goals for this project was to fabricate ceramic to thicknesses

of between 150-300 µm that had tolerances which were equivalent to or better than industry standards. Piezoelectric elements were sintered to greater than

97%  of  theoretical  density  and  had  grain  sizes

between 2 and 10 µm. The bulk density and grain size  markedly  influenced  the  piezoelectric properties of the sintered ceramic [22]. An outcome of this project was the development of some intellectual property centred on a process for shaping piezoelectric ceramic and fabricating arrays from taped material [23]. Distortion-free piezoelectric tapes and laminates were prepared forming a variety of shapes such as disks with curved surfaces and flat rectangular elements used to form tiers in piezoelectric/polymer composites. Figure 3 below illustrates some of the piezoelectric elements fabricated from green tape.


The specification of the piezoelectric ceramic developed was based on a Hans Ruckert formulation, and is listed in table 3 below along with a comparison of with ceramics used by Ausonics which corresponded closely with a modified lead titanate. The development of this ceramic tape technology showed that cost was not an impediment to traditional machining ceramic or dice  and  fill  composite  arrays.  However,  in  the early 1990s under the direction of new management (Pacific Dunlop Ltd) the development of this technology ceased.




Hans Ruckerts 1970 patent has been the backbone

of the materials produced over the past 40 years. Materials based on the guidelines expounded in this

patent are still in manufacture. However, the two

materials that are currently in largest volume production are not based on the patent.

Over the last forty years, the Thales group has had

to  build  transducers  and  supply  ceramic components which have been built? to print. As

the  designs  invariably  come  from  the  USA  or

Europe, they have specified ceramics compliant to MIL_STD_1376 (SHIPS) 1970 or its later editions such as MIL-STD-1376B (SH). On a number of occasions, the customer has only been willing to accept full material compliance.


As an example, we can consider the properties of the Thales TLZ 1 material made in accordance with the Plessey Ducon Australian patent. This material

exceeds  the  specification  on  some  batches.  The

material also changes its capacitance under uniaxial pressure a little more than typical lots of a more

traditional type I material.


Because of such difficulties, it has been necessary to produce materials which are totally compliant with MIL-STD 1376B. Yet even this does not satisfy all requirements. It is well known that PZT materials are sensitive to small amounts of dopants, to the raw materials used and to the processing steps. When all these are optimised, materials with enhanced properties can be produced. An example of this is the TLZ 3 material which is currently produced in the largest volumes.


With particular attention paid to the above three factors, the properties of many batches of this material would strictly fail the MIL-STD 1376B criteria for coupling and d33. For most customers this is a bonus”. A side effect of this attention to composition and processing has been a material which also a significantly improved stability with respect to voltage and pressure. The TLZ 2 formulation has been similarly enhanced over the standard type II material.


Another ceramic formulation which is in large volume production at this current time is a modified lead titanate, TL 1. The principal dopant is calcium titanate but there are numerous minor additives which are designed to improve the ability to pole this material, lower its tan δ and to increase its stability with time. This material is similar to compositions used by other suppliers but some key additives have been identified which are critical to achieving the benefits noted above. Use of such unique  compositions  combined  with  tight processing  controls  over  green  and  sintered densities has significant benefits.   In a recent competitive comparison, a Thales customer praised the   uniformity  and   consistency  of   the   TL   1 dielectric constant and d33  which enabled them to reduce the system tolerances by 50%.


Further improvements to compositions inspired by the 1970 patent are still being undertaken in current piezoelectric formulations. In 2012, the TLZ 5 composition and processing were modified to produce a material with a d33 > 900 pC/N. Potential modifications to the high manganese doped materials have been identified which may enable materials with voltage stabilities similar to type III to be produced with d33 values similar to type I. Lastly, materials such as TLZ 4 have demonstrated the ability to combine the stability of a type I with the piezoelectric properties of a type II. Modifications are being trialled which could produce a voltage stable type V equivalent.




This paper has reviewed the evolution of electronic

ceramics in Australia over the past 50 years with specific reference to the development of dielectrics

and piezoelectric devices in targeted applications.

Although the focus of this review did not touch on research  underpinning  the  commercial  products,

there are  several groups across academia and  at

national labs which continue to pursue developing materials and         process  technologies.        These

developments will ensure that local expertise and

manufacturing capabilities remain an integral part of our niche ceramic manufacturing hub.




1.      Mellor,   D.,   Australian   War   Memorial official history ww2_civil_vol5_. In: s.l.: Australian War Memorial, (1958), Chapter

21 Communications.

2.      SMH,   13   July   1932.   Sydney   Morning


3.      SMH,   23   July   1963.   Sydney   Morning


4., n.d.

5.      SMH,   24   June   1949.   Sydney  Morning


6.      SMH,   4   May   1954.   Sydney   Morning


7.     SMH, 14 Oct 1954. Sydney Morning Herald

8.      Gleeson,  J.  M.,  1946.  Steatite  for  High

Frequency Insulation. J. Br. I.R.E. Vol [6], pp. 20-32.

9.      Randall,  C.,  2009.  History  of  the  First Ferroelectric Oxide, BaTiO3, s.l.: American Ceramics Society.

10. Thales    Austalia    Internal    Report GCM04572, n.d. GCM04572 Ruckert report History of Ceramic Capacitor Development in Australia, s.l.: s.n.

11.   Thales       Australia       Internal       report

GCM02597, n.d. GCM02597 Ruckert report

266  Survey  on  Ceramic  Capacitors  1971, s.l.: s.n.

12. Thales   Australia   Internal   Report GCM02599,  n.d.  GCM02599  Ruckert Report  272  on  NPO  Dielectrics Formulations and History, s.l.: s.n.

13. Thales   Australia   Internal   Report GCM02604,  n.d.  GCM02604  Ruckert Report  275  on  Development  of  Type  2

Ceramic Dielectrics 1973, s.l.: s.n.

14. Thales   Australia   Internal   Report GCM02011, n.d. GCM02011 Hans Ruckert Report on Design of Barrier Layer Capacitors, s.l.: s.n.

15. Ruckert,   H.   F.,   Development   of Piezoeelctric Ceramics in Australia. J. Aust. Cer Soc, Vol [8], (1972), 33-37.

16.   Mil-STD-1376B       (SH),        Piezoelectric

Ceramic      Material     and     Measurement

Guidelines   for   Sonar   Transducers,   US Department of Defence 1995

17.   Hiromi  O  et  al.,  Matsushita,  US  patent

3268453 (1966)].

18.   Ruckert,  H.,  1970.  Australia,  Patent  No.


19.  Griffiths, K., 2004. An historical look at ultrasound as  an  Australian innovation on the occasion of the ultrasound stamp issued by Australia Post -18th May 2004. ASUM Ultrasound Bulletin, August, pp. 22-26.

20.   Woolfrey J. L., The feasibility study on the fabrication      of      piezoceramic/polymeric

transducer arrays, ANSTO report 1987.

21.   Mistler R. E. and Twiname   E. R. Tape

Casting:    Theory    and    Practice,     2000

Published by the American Ceramic Society, Westerville OH

22.   Patel N. D and Nicholson P. S. Am. Ceram.

Soc. Bull 65 (1986) 783-79.

23.   Aust patent filing “Tape Casting, Shaping and   Array   Fabrication   of   Piezoelectric

Ceramic Nov 1989.